Investigating Representative Elementary Volume of a fracture network from outcrop data to drive Discrete Fracture Network modeling

@inproceedings{MartinelliRM2019,
 abstract = { Fracture networks in hydrocarbon reservoirs are complex structures that are difficult to characterize due to the incompleteness and the uncertainties of data obtained from borehole and seismic analysis. The use of outcrop analogues can help to better parametrize fracture networks and their organization. However, the Discrete Fracture Network (DFN) modeling at the reservoir scale remain difficult due to the impossibility to explicitly represent every fractures through all scale. This issue can be partially solved integrating the effect of small scale fractures as a property is a Representative Elementary Volume (REV). In this work, we focus on uniform, randomly and poorly organized fracture networks. Based on that assumption, we investigate the size of the Representative Elementary Volume (REV) of fracture density and intensity (P20, P21) using square scan area of various size on outcrop data of a fracture network composed from three main fracture sets from the Island of Gozo (Maltese Islands). We demonstrated that for every fracture set a REV exist for all the investigated parameters. We verified that the simulation of stochastic Discrete Fracture Networks, driven by the parameters derived from the field, generates models which the REV match the one computed from outcrop. At the end several DFN models at the REV scale were realized and for each of them both P20 and P21 values were computed to obtain a P20 and P21 distribution histogram that were used to populate the 3D model of the Gozo Island. },
 author = { Martinelli, Mattia AND Bistacchi, Andrea AND Bonneau, Francois AND Caumon, Guillaume AND Meda, Marco },
 booktitle = { 2019 Ring Meeting },
 publisher = { ASGA },
 title = { Investigating Representative Elementary Volume of a fracture network from outcrop data to drive Discrete Fracture Network modeling },
 year = { 2019 }
}